Skip Navigation LinksHome > Blogs > The Spine Blog > Epidural Steroid Injections for Spinal Stenosis: Still No Ev...
The Spine Blog
Friday, February 15, 2013
Epidural Steroid Injections for Spinal Stenosis: Still No Evidence of Effectiveness

The role of epidural steroid injection (ESI) in spinal stenosis (SpS) has been controversial, with no Level 1 or Level 2 evidence clearly demonstrating effectiveness in this population. Despite the lack of evidence for this procedure, it is generally viewed as first line treatment for SpS, especially for patients who prefer to avoid surgery. Most studies of ESI have included mixed populations (i.e. disk herniation,  stenosis, degenerative spondylolisthesis) and frequently have compared steroid to a local anesthetic or saline control, a potential problem as the local anesthetic or saline could have a therapeutic benefit. The Spine Patient Outcomes Research Trial (SPORT) provides a large observational database with which subgroup analyses can be performed. Dr. Radcliff and his colleagues from Thomas Jefferson University and Dartmouth performed such an analysis, comparing surgical and non-operative outcomes between patients who received an ESI within 3 months of enrollment in SPORT and those who had never received ESI. Patients who received an ESI prior to enrollment or after 3 months from enrollment were excluded in order create a more homogenous ESI population. Over the four years of follow-up, they found significantly worse surgical and non-operative outcomes in the ESI patients on some outcome measures (primarily the SF-36 physical function and bodily pain scales), while there were no significant differences on the Oswestry Disability Index (ODI) or the Sciatica Bothersomeness Index. Epidural steroid injection did not clearly reduce surgery rates, with increased crossover to surgery among ESI patients who initially chose or were assigned to non-operative treatment and increased crossover to non-operative treatment among ESI patients who initially chose or were assigned to surgery. Somewhat surprisingly, ESI patients had significantly longer surgery times and length of hospital stays after surgery. Additionally, ESI patients had surgery complicated by dural tear at over twice the rate compared to non-ESI patients (15% vs. 7%), though this difference was not significant (p=0.21).


 Based on the results, the authors suggest that ESI is associated with worse surgical and non-operative results for SpS patients. While this was true on some outcome measures, other measures—such as the ODI—did not show any significant differences. This study was not an RCT comparing outcomes between patients randomized to ESI and no ESI, so the results could be affected by unmeasured confounders. In other words, there could be factors associated with choosing an ESI that are responsible for the worse outcomes rather than the ESI itself. As such, these data may not convincingly demonstrate that ESI causes worse outcomes, however, there is no suggestion that ESI was beneficial in this population. One concerning finding was that patients who had underwent ESI had significantly longer OR times and lengths of stay. Additionally, the durotomy rate was twice as high in the ESI group, though this difference was not significant due to the small numbers involved. While there may have been some confounding by the ESI patients undergoing fusion at a slightly higher rate (15% vs. 10%), these operative factors suggest that ESI could create adhesions or have other effects that make surgery more difficult. While this study is not going to mean the end of ESI for SpS patients, it adds to a flawed but extensive literature suggesting that ESI is not very effective for this condition. There is very little evidence that any non-operative treatment is effective for SpS, so ESI will likely remain in our armamentarium as an option for patients who wish to avoid surgery or who have medical comorbidities that make surgery risky. This study does provide us with good data to share with our SpS patients who are considering their treatment options, and they should be informed that ESI is generally not helpful in the long-term.


Please read Dr. Radcliff’s article on this topic in the February 15 issue. Will this article change how you use ESI in the SpS population? Let us know by posting a comment on The Spine Blog.


Adam Pearson, MD, MS

Associate Web Editor

Dr. Snooze DrSnooze said:
I'm really enjoying your website, nice information you've shared. Thanks!
Dr. David M. Glick said:
( 3of 3) By no mean is this meant to devalue the author’s hard work. It is clear that much effort was dedicated to the execution study and evaluation of the clinical data. It is time though for someone to raise the bar and start being more patient and pathology specific when it comes to studies to better help assess effective treatments for back pain, including spinal stenosis.
Dr. David M. Glick said:
(2 of 3) The bottom line is this study is no more flawed than most all that are focus toward back pain. For anyone who is truly interested in addressing one of the larger underlying causes of pain and disability that effects the population, it will be necessary to take a more individualized, patient oriented approach. With more precise differential diagnosis, including thorough hands-on examination by physicians who truly understand the clinical aspects of each of the different potential pathologies and pain generators in the back, which often overlap complicating matter further. Along with electrodiagnostic studies tailored to better assess the patient’s pathology, imaging studies should be recognized as valued as a piece of the puzzle necessary to view the entire clinical picture rather than the primary diagnosis criteria, especially since the practice of relying upon the imaging study for the assessment of back pain has never been clinically validated.
Dr. David M. Glick said:
(1 of 3) With all due respect, there are major flaws to the study. Unfortunately it does exemplify the problem we have in dealing with back pain in general. Simply relying upon the results of imaging studies and patient complaints to render a diagnosis is at best questionable. Literature has demonstrated a near 50% incidence or presence of the same abnormal findings in asymptomatic patients. And technically, in the case where the ESI was effective, a situation was created where such pathologies are present on imagining studies, but are no longer resulting in clinical symptoms. Furthermore, there are several variations to the theme for ESI’s. Different pathologies often require and respond differently to each of the different approaches for and ESI.
About the Blog

Spine Journal
This Blog provides a forum for discussion about high impact articles published in Spine, including the bi-annual publication of "Evidenced-Based Recommendations for Spine Surgery." Website users can use this forum to discuss how the articles have affected their practice and query the authors about their findings and recommendations.